import the essential packages¶
In [1]:
import numpy as np
import cv2
import glob
import os
import matplotlib.image as mpimg
import pickle
import matplotlib.pyplot as plt
First, I'll compute the camera calibration using chessboard images¶
In [2]:
images = glob.glob('camera_cal/calibration*.jpg')
imagess = []
for i in range(len(os.listdir('./camera_cal'))):
imagess.append(mpimg.imread('./camera_cal' + '/' + os.listdir('./camera_cal')[i]))
objp = np.zeros((6*9,3), np.float32)
objp[:,:2] = np.mgrid[0:9,0:6].T.reshape(-1,2)
# Arrays to store object points and image points from all the images.
output_calibration =[]
objpoints = [] # 3d points in real world space
imgpoints = [] # 2d points in image plane.
# Step through the list and search for chessboard corners
for fname in images:
img = cv2.imread(fname)
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
# Find the chessboard corners
ret, corners = cv2.findChessboardCorners(gray, (9,6),None)
# If found, add object points, image points
if ret == True:
objpoints.append(objp)
imgpoints.append(corners)
output_calibration.append(cv2.drawChessboardCorners(img, (9,6), corners, ret))
cv2.destroyAllWindows()
visualize camera calibration¶
In [3]:
for i in range(len(output_calibration)):
plt.imshow(output_calibration[i])
plt.show()
visualize camera calibration¶
In [4]:
for i in range(len(imagess)):
img_size = (imagess[i].shape[1], imagess[i].shape[0])
ret, mtx, dist, rvecs, tvecs = cv2.calibrateCamera(objpoints, imgpoints, img_size,None,None)
dst = cv2.undistort(imagess[i], mtx, dist, None, mtx)
plt.imshow(dst)
plt.show()
Apply Bird view transform on images¶
In [5]:
images_warp=[]
for i in range(len(os.listdir('./test_images'))):
images_warp.append(mpimg.imread('./test_images' + '/' + os.listdir('./test_images')[i]))
for i in range(len(images_warp)):
img_size = (images_warp[i].shape[1], images_warp[i].shape[0])
ret, mtx, dist, rvecs, tvecs = cv2.calibrateCamera(objpoints, imgpoints, img_size,None,None)
images_warp[i] = cv2.undistort(images_warp[i], mtx, dist, None, mtx)
src = np.float32([
[220, 720],
[1150, 720],
[570, 470],
[720, 470]
])
dst = np.float32([
[320, 720],
[920, 720],
[320, 0],
[920, 0]
])
bird_images = []
for img in images_warp:
img_size = (img.shape[1], img.shape[0])
transform = cv2.getPerspectiveTransform(src,dst)
bird_img = cv2.warpPerspective(img, transform, img_size)
bird_images.append(bird_img)
plt.imshow(img)
plt.show()
print("Warped image")
plt.imshow(bird_img, cmap='gray')
plt.show()
Apply Sobel transform on images¶
In [6]:
for image in images_warp:
gray = (cv2.cvtColor(image, cv2.COLOR_RGB2Lab))[:,:,0]
sobel = cv2.Sobel(gray, cv2.CV_64F, 0, 1)
scaled_sobel = np.uint8(255*np.absolute(sobel)/np.max(np.absolute(sobel)))
img_bin = np.zeros_like(scaled_sobel)
img_bin[(scaled_sobel >= 35) & (scaled_sobel <= 230)] = 1
plt.imshow(image)
plt.show()
plt.imshow(img_bin, cmap='gray')
plt.show()
Apply Sobel transform on images¶
In [7]:
sob_out = []
for image in bird_images:
gray = (cv2.cvtColor(image, cv2.COLOR_RGB2Lab))[:,:,0]
sobel_out = cv2.Sobel(gray, cv2.CV_64F, 1, 0)
scaled_sobel = np.uint8(255*np.absolute(sobel_out)/np.max(np.absolute(sobel_out)))
img_bin = np.zeros_like(scaled_sobel)
img_bin[(scaled_sobel >= 25) & (scaled_sobel <= 150)] =1
sob_out.append(img_bin)
plt.imshow(image, cmap='gray')
plt.show()
plt.imshow(img_bin, cmap='gray')
plt.show()
Apply color filter on images¶
In [8]:
hls_out = []
for image in bird_images:
hls = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)[:, :, 2]
hls_bin = np.zeros_like(hls)
hls_bin[(hls >= 125) & (hls <= 255)] = 1
hls_out.append(hls_bin)
plt.imshow(image)
plt.show()
plt.imshow(hls_bin, cmap='gray')
plt.show()
Apply color filter on images¶
In [9]:
for image in images_warp:
hls = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)[:, :, 2]
hls_bin = np.zeros_like(hls)
hls_bin[(hls >= 160) & (hls <= 255)] = 1
plt.imshow(image)
plt.show()
plt.imshow(hls_bin, cmap='gray')
plt.show()
apply both sobel and color filter on images¶
In [11]:
out_img = []
for i in range(len(hls_out)):
img_out = np.zeros_like(sob_out[i])
img_out[( hls_out[i] == 1) | (sob_out[i] == 1)] = 1
plt.imshow(images_warp[i])
out_img.append(img_out)
plt.show()
plt.imshow(img_out, cmap='gray')
plt.show()
In [16]:
def weighted_img(img, initial_img, α=0.8, β=1., γ=0.):
return cv2.addWeighted(initial_img, α, img, β, γ)
changes = []
for i in range(len(out_img)):
changes.append(weighted_img(out_img[i], images_warp[i]))
In [ ]: